ABSTRACT
Using a large Web search service as a case study, we highlight the challenges that modern Web services face in understanding and diagnosing the response time experienced by users. We show that search response time (SRT) varies widely over time and also exhibits counter-intuitive behavior. It is actually higher during off-peak hours, when the query load is lower, than during peak hours. To resolve this paradox and explain SRT variations in general, we develop an analysis framework that separates systemic variations due to periodic changes in service usage and anomalous variations due to unanticipated events such as failures and denial-of-service attacks. We find that systemic SRT variations are primarily caused by systemic changes in aggregate network characteristics, nature of user queries, and browser types. For instance, one reason for higher SRTs during off-peak hours is that during those hours a greater fraction of queries come from slower, mainly-residential networks. We also develop a technique that, by factoring out the impact of such variations, robustly detects and diagnoses performance anomalies in SRT. Deployment experience shows that our technique detects three times more true (operator-verified) anomalies than existing techniques.
- Central limit theorem. http://en.wikipedia.org/wiki/Central_limit_theorem.Google Scholar
- Measure page load time. https://developers.google.com/speed/docs/pss/LatencyMeasure.Google Scholar
- Pagespeed tools family. https://developers.google.com/speed/pagespeed/.Google Scholar
- WebPageTest - website performance and optimization test. http://www.webpagetest.org/.Google Scholar
- Yslow. http://yslow.org/.Google Scholar
- T. Abdelzaher, K. Shin, and N. Bhatti. Performance guarantees for Web server end-systems: A control-theoretical approach. IEEE Transactions on Parallel and Distributed Systems, 2002. Google ScholarDigital Library
- E. Agichtein, E. Brill, S. Dumais, and R. Ragno. Learning user interaction models for predicting Web search result preferences. In ACM SIGIR, 2006. Google ScholarDigital Library
- M. Al-Fares, K. Elmeleegy, B. Reed, and I. Gashinsky. Overclocking the Yahoo! CDN for faster Web page loads. In IMC, 2011. Google ScholarDigital Library
- P. Bahl, R. Chandra, A. Greenberg, S. Kandula, D. Maltz, and M. Zhang. Towards highly reliable enterprise network services via inference of multi-level dependencies. In ACM SIGCOMM, 2007. Google ScholarDigital Library
- L. Barroso, J. Dean, and U. Holzle. Web search for a planet: The Google cluster architecture. IEEE Micro, 2003. Google ScholarDigital Library
- M. Butkiewicz, H. Madhyastha, and V. Sekar. Understanding website complexity: Measurements, metrics, and implications. In IMC, 2011. Google ScholarDigital Library
- Y. Chen, S. Jain, V. Adhikari, and Z. Zhang. Characterizing roles of front-end servers in end-to-end performance of dynamic content distribution. In IMC, 2011. Google ScholarDigital Library
- J. Dean. Achieving rapid response times in large online services. In Berkeley AMPLab Cloud Seminar, 2012.Google Scholar
- M. Freedman, E. Freudenthal, and D. Mazieres. Democratizing content publication with coral. In USENIX NSDI, 2004. Google ScholarDigital Library
- A. Gelman. Analysis of variance, 2005. http://www.stat.columbia.edu/ gelman/research/unpublished/econanova.pdf%.Google Scholar
- C. Gkantsidis and P. Rodriguez. Network coding for large scale content distribution. In IEEE INFOCOM, 2005.Google ScholarCross Ref
- C. Huang, A. Wang, J. Li, and K. Ross. Measuring and evaluating large-scale cdns. In IMC, 2008. Google Scholar
- H. Jiang, J. Li, Z. Li, and X. Bai. Performance evaluation of content distribution in hybrid CDN-P2P network. In IEEE Future Generation Communication and Networking, 2008. Google ScholarDigital Library
- K. Johnson, J. Carr, M. Day, and M. Kaashoek. The measured performance of content distribution networks. Elsevier, 2001.Google ScholarDigital Library
- S. Kandula, R. Mahajan, P. Verkaik, S. Agarwal, J. Padhye, and P. Bahl. Detailed diagnosis in enterprise networks. In ACM SIGCOMM, 2009. Google ScholarDigital Library
- J. Kangasharju, K. Ross, and J. Roberts. Performance evaluation of redirection schemes in content distribution networks. Elsevier, 2001.Google ScholarDigital Library
- R. Kohavi, R. M. Henne, and D. Sommerfield. Practical guide to controlled experiments on the Web: Listen to your customers not to the HiPPO. In SIGKDD, 2007. Google ScholarDigital Library
- R. Krishnan, H. Madhyastha, S. Srinivasan, S. Jain, A. Krishnamurthy, T. Anderson, and J. Gao. Moving beyond end-to-end path information to optimize CDN performance. In IMC, 2009. Google ScholarDigital Library
- A. Pinto. How to measure page load time with google analytics. http://www.yottaa.com/blog/bid/215491/How-to-Measure-Page-Load-Time-Wit%h-Google-Analytics.Google Scholar
- D. G. Rees. Foundations of statistics, volume 214. Chapman & Hall, 1987.Google Scholar
- M. Roughan, A. Greenberg, C. Kalmanek, M. Rumsewicz, J. Yates, and Y. Zhang. Experience in measuring backbone traffic variability: Models, metrics, measurements and meaning. In ACM IMW, 2002. Google ScholarDigital Library
- A. Singhal and M. Cutts. Official Google webmaster central blog: Using site speed in Web search ranking, 2009. http://googlewebmastercentral.blogspot.com/2010/04/using-site-speed-in-%web-search-ranking.html.Google Scholar
- Y. Song, V. Ramasubramanian, and E. Sirer. Optimal resource utilization in content distribution networks. Technical Report TR2005--2004, Cornell University, 2005.Google Scholar
- S. Stefanov. Progressive rendering via multiple flushes. http://www.phpied.com/progressive-rendering-via-multiple-flushes/.Google Scholar
- W. Taylor. Change-point analysis: a powerful new tool for detecting changes, 2000. http://www.variation.com/cpa/tech/changepoint.html.Google Scholar
- S. Triukose, Z. Wen, and M. Rabinovich. Measuring a commercial content delivery network. In ACM WWW, 2011. Google ScholarDigital Library
- X. S. Wang, A. Balasubramanian, A. Krishnamurthy, and D. Wetherall. Demystify page load performance with WProf. In USENIX NSDI, 2013. Google ScholarDigital Library
- R. White and S. Dumais. Characterizing and predicting search engine switching behavior. In ACM conference on Information and knowledge management, 2009. Google ScholarDigital Library
- S. Zander, L. Andrew, G. Armitage, G. Huston, and G. Michaelson. Mitigating sampling error when measuring Internet client IPv6 capabilities. In IMC, 2012. Google ScholarDigital Library
Index Terms
- A provider-side view of web search response time
Recommendations
A provider-side view of web search response time
Using a large Web search service as a case study, we highlight the challenges that modern Web services face in understanding and diagnosing the response time experienced by users. We show that search response time (SRT) varies widely over time and also ...
Using HMM for predicting response time of web services
CUBE '12: Proceedings of the CUBE International Information Technology ConferenceAvailability of services on Internet has provided unique opportunity to customers as well as providers for conducting e-business. This new business paradigm can succeed provided selection of services is accomplished to customer satisfaction in terms of ...
Composing Web Services: A QoS View
An Internet application can invoke several services--a stock-trading Web service, for example, could invoke a payment service, which could then invoke an authentication service. Such a scenario is called a composite Web service, and it can be specified ...
Comments